Research On Load Balancing Control Of Distributed Storage System For Aerospace Telemetry Data

Telemetry system in the field of Aeronautics and Astronautics undertakes the responsibility of monitoring,transmitting and processing the operation status and data of the launch vehicle.Telemetry data provides data support for ground workers to monitor and evaluate the operation status of the launch vehicle.It is an important basis for ensuring the safety and reliability of the launch vehicle.In order to meet the requirements of real-time monitoring,distributed storage system is often used to store data.However,due to the large amount of telemetry data and many types of data,distributed storage system is prone to low read-write efficiency and stability decline,which can not meet the basic requirements of real-time status monitoring and security evaluation of launch vehicles.Aiming at the reliability and efficiency of distributed storage system in aerospace applications,this paper elaborates the important role of load balancing technology in detail.Then,aiming at the deficiencies of existing load balancing algorithms in request allocation and data storage statics,the following research is carried out:In the process of telemetry data storage,the existing load balancing algorithms do not consider the real-time load status of each node in the system and the difference of different requests.This thesis establishes a data heat model based on the node status and data attributes,and uses the heat value to reflect the impact of reading and writing the data on the node load.Long-term and short-term memory network is used to predict data heat to solve the problem of time delay.Finally,load clustering is used to guide requests to be allocated to low-load nodes.For the problem of load control failure due to the static nature of data storage during the reading of telemetry data,this thesis establishes a data scheduling model based on the heat value,and simulates the heat flow to dispatch the high heat data of the high load node to the low load node to balance the load.At the same time,the optimal scheduling scheme is searched by genetic algorithm,so that the impact of the scheduling process on the system load is as small as possible.Combined with data heat model and data scheduling model,CloudSim is used to build a distributed storage simulation platform,and then simulate the read and write operations of telemetry data in real space missions to realize load balancing control of distributed storage systems and compare with other load balancing algorithms.Analyze and verify the effectiveness and applicability of the method studied in this paper to system load control,and finally ensure the stability of the system and the efficiency of reading and writing,and meet the basic requirements of telemetry data in analytical applications.